摘要
文章利用水质综合指数模型进行场景创新,结合管道内水质物理环境稳定的特点,基于集成学习算法,开发了管道水质模型PWQI(Pipeline Water Quality Index)。该模型选取17项管道水质参数并分为5类。为避免水质参数赋权的单一性,采用专家意见与熵权法进行水质参数组合赋权。采用CatBoost算法进行回归预测,与传统水质评估方法对比,结果表明PWQI模型对管道水质状况的预测效果最为显著,能准确反映管道水质状况,为管道水质评价提供了依据。
In this paper,the water quality comprehensive index model is used for scenario innovation.Combined with the characteristics of stable physical environment of water quality in the pipeline,the pipeline water quality model PWQI is developed based on the integrated learning algorithm.The model selects 17 pipeline water quality parameters and divides them into 5 categories.In order to avoid the single weighting of water quality parameters,the combination weighting of water quality parameters is carried out by expert opinion and entropy weighting method.The CatBoost algorithm is used for regression prediction.Compared with the traditional water quality assessment method,the results show that the PWQI model has the most notable prediction effect on the pipeline water quality,which can accurately reflect the water quality of the pipeline and provide a basis for pipeline water quality assessment.
作者
聂立俊
NIE Lijun(School of Information Engineering,North China University of Water Resources and Electric Power,Zhengzhou 450046,China)
出处
《现代信息科技》
2025年第10期45-49,共5页
Modern Information Technology
关键词
水质指数
集成算法
组合赋权
水质评估
water quality index
integrated algorithm
combined weighting
water quality assessment
